Scientific Publications - Work Done by Microbiology Reader Bioscreen C

Free Online Full-text Article

Eur. J. Biochem. 262, 191-201 (1999)

Autoregulation of yeast pyruvate decarboxylase  gene expression requires  the enzyme but not  its catalytic activity

Ines Eberhardt^1,2, Håkan Cederberg³, Haijuan Li⁴, Stephan König², Frank Jordan⁴ and Stefan Hohmann^1,3,5

¹ Laboratorium voor Moleculaire Celbiologie, Katholieke Universiteit Leuven, Flanders, Belgium; ² Institut für Biochemie, Fachbereich Biochemie/Biotechnologie, Martin-Luther-Universität Halle-Wittenberg, Germany; ³ Institut für Mikrobiologie und Genetik, Technische Universität Darmstadt, Germany; ⁴ Department of Chemistry, Rutgers University, Newark, New Jersey, USA; ⁵ Department of Cell and Molecular Biology/Microbiology, Göteborg University, Sweden

Correspondence to S. Hohmann, Department of Cell and Molecular Biology/Microbiology, Göteborg University, Box 462, SE-405 30 Göteborg, Sweden. Fax: + 46 31 7732599. Tel: + 46 31 7732595. E-mail: hohmann@gmm.gu.se

 
In the yeast, Saccharomyces cerevisiae, pyruvate decarboxylase (Pdc) is encoded by the two isogenes PDC1 and PDC5. Deletion of the more strongly expressed PDC1 gene stimulates the promoter activity of both PDC1 and PDC5, a phenomenon called Pdc autoregulation. Hence, pdc1 strains have high Pdc specific activity and can grow on glucose medium. In this work we have characterized the mutant alleles pdc1-8 and pdc1-14, which cause strongly diminished Pdc activity and an inability to grow on glucose. Both mutant alleles are expressed as detectable proteins, each of which differs from the wild-type by a single amino acid. The cloned pdc1-8 and pdc1-14 alleles, as well as the in-vitro-generated pdc1-51 (Glu51Ala) allele, repressed expression of PDC5 and diminished Pdc specific activity. Thus, the repressive effect of Pdc1p on PDC5 expression seems to be independent of its catalytic activity. A pdc1-8 mutant was used to isolate spontaneous suppressor mutations, which allowed expression of PDC5. All three mutants characterized had additional mutations within the pdc1-8 allele. Two of these mutations resulted in a premature translational stop conferring phenotypes virtually indistinguishable from those of a pdc1 mutation. The third mutation, pdc1-803, led to a deletion of two amino acids adjacent to the pdc1-8 mutation. The alleles pdc1-8 and pdc1-803 were expressed in Escherichia coli and purified to homogeneity. In the crude extract, both proteins had 10% residual activity, which was lost during purification, probably due to dissociation of the cofactor thiamin diphosphate (ThDP). The defect in pdc1-8 (Asp291Asn) and the two amino acids deleted in pdc1-803 (Ser296 and Phe297) are located within a flexible loop in the domain. This domain appears to determine the relative orientation of the and domains, which bind ThDP. Alterations in this loop may also affect the conformational change upon substrate binding. The mutation in pdc1-14 (Ser455Phe) is located within the ThDP fold and is likely to affect binding and/or orientation of the cofactor in the protein. We suggest that autoregulation is triggered by a certain conformation of Pdc1p and that the mutations in pdc1-8 and pdc1-14 may lock Pdc1p in vivo in a conformational state which leads to repression of PDC5.

Keywords: gene expression; glycolysis; pyruvate decarboxylase; thiamin diphosphate; yeast.

Abbreviations: PDC, pyruvate decarboxylase gene; Pdc, pyruvate decarboxylase; ACT1, actin gene; ThDP, thiamin diphosphate; ORF, open reading frame; YPD, yeast extract peptone medium containing 2% glucose; YPE, yeast extract peptone medium containing 3% ethanol; SD, defined synthetic medium; SDtrp, SD lacking trptophan; ADH, alcohol dehydrogenase.

 
Pyruvate decarboxylase (Pdc) catalyses the degradation of the end product of glycolysis, pyruvate, to acetaldehyde and CO₂. In the yeast Saccharomyces cerevisiae two structural genes, PDC1 and PDC5, encode the isoenzymes Pdc1p and Pdc5p [1-4]. Pdc1p and Pdc5p are 88% identical, closely related over the entire sequence to Pdc from other organisms, and both independently form active enzymes [3,5-7]. A third structural gene, PDC6, has been characterized in S. cerevisiae [5]. Pdc6p is an active Pdc [6,8,9], but is apparently not involved in glucose fermentation [5].

In actively fermenting yeast cells, PDC1 is strongly expressed [1,3,10-12]. Under the same conditions, expression of PDC5 is hardly detectable [3,12]. Accordingly, deletion of PDC5 does not noticeably reduce the Pdc specific activity [3,5]. Remarkably, in pdc1 mutants, the Pdc specific activity is as high as 50-80% of that of the wild-type: this effect is due to strongly enhanced expression of PDC5. As the promoter activity of PDC1 is also stimulated in a pdc1 strain it appears that expression of PDC1 and PDC5 is controlled by an autoregulation mechanism [2,3,11,13]. Pdc1 pdc5 double mutants have no detectable Pdc activity and are unable to grow on glucose as the sole carbon source; this is because their respiration capacity is insufficient to support sugar catabolism. In addition, a proportion of the acetaldehyde that is produced in the Pdc reaction is required for acetyl-CoA production in the cytosol for biosynthetic pathways [3,4,14-16].

The mechanisms that mediate the stimulation of PDC5 (and PDC1) promoter activity in a pdc1 strain are not known. Liesen et al. identified a promoter element termed ERA which they believed mediates repression of PDC1 and PDC5 in the presence of the wild-type PDC1 gene, and at the same time is required for the control of PDC1 expression by the carbon source [11]. However, deletion analysis of the PDC5 promoter did not support the function of this element [16a].

The structure-function relationship of Pdc has been studied intensively and the three-dimensional structure of the protein has been resolved from two different types of crystals [17-19]. The protein is a homotetramer, which readily dissociates into very stable dimers: each subunit consists of three domains of approximately equal size: , and . The and domains have similar structural organization. The holoenzyme binds four molecules of thiamin diphosphate (ThDP) and four Mg²⁺ ions as cofactors. Within each dimer the two monomers both contribute with their and domains to the binding of each molecule of ThDP. Pdc is activated by the substrate pyruvate (or its analogues) via an allosteric site and the pathway of activation to the catalytic centre has been mapped [6,18,20,21]. When Pdc was crystallized in the presence of the substrate analogue pyruvamide, fundamental differences in the tetramer arrangement were observed, suggesting a major conformational change upon substrate binding [22].

In this work we have investigated the role of Pdc1p in the control of PDC5 expression. Well before it was known that Pdc is encoded by isoenzymes in S. cerevisiae, Schmitt and Zimmermann [23] had isolated mutant alleles of PDC1 which expressed very low Pdc activity, accumulated pyruvate and were unable to grow on glucose. Because the phenotypes of these mutants appeared to contrast sharply with those of a pdc1 mutant [2,3,13], we were interested in the molecular nature of these mutations and their effect on expression of PDC5. We confirm that these alleles encode catalytically defective Pdc and we show that they still cause repression of PDC5. Thus it appears that the autoregulation of PDC gene expression is not directly related to the catalytic activity of Pdc but to other properties of the protein.

 
Yeast strains
The yeast strains used in this work are summarized in Table 1. Original pdc1-8 and pdc1-14 mutants, described by Schmitt and Zimmermann [23], were crossed with YSH 6.36.-3B, a haploid derivative of the diploid M5 [13], to yield the pdc1-8 and pdc1-14 strains used in this study.

 
Three different deletion alleles of PDC1 were used. Pdc1-1 is a deletion of part of the open reading frame (ORF) from position +107 to +1378 [13], pdc1-2 is a complete deletion of the ORF and pdc1-3 is a complete deletion of the ORF plus the upstream regulatory sequences up to position -801 relative to the translational start side. These two deletions, as well as the complete deletion of the PDC5 ORF in pdc5-2 strains, were constructed by PCR [24]. The pdc5-1 mutation, which encompasses a region from +105 relative to the translational start until 392 bp downstream of the stop codon of PDC5, has been described previously [3]. The PDC5 promoter-lacZ reporter construct was described by Hohmann [12]. Standard yeast genetic techniques were used for crossing of yeast strains and for their cultivation [25].

Growth conditions
Growth properties were tested on yeast extract peptone medium containing either 2% glucose (YPD) or 3% ethanol (YPE). For growth curves, yeast cells were grown for 3 days in defined synthetic medium (SD medium [25]) lacking Trp (SDtrp) and with 3% ethanol as a carbon source. These cultures were diluted to a D_{600 nm} of 0.2 into fresh SDtrp medium with 2% glucose as a carbon source. Growth was monitored in microtiter plates using the Bioscreen C system (Labsystems).

For the determination of enzyme activities and for Western blot analysis, yeast cells were pregrown for 2 days in YPE (noninducing conditions) and then 1 mL of this culture was inoculated into 6 mL YPD containing 8% glucose to induce PDC gene expression. After 6 h of vigorous shaking at 30 °C, cells were harvested by centrifugation. For Northern blot analysis, cells were harvested either before or 1 h after glucose addition. Yeast transformants were grown in SD instead of YP medium [25] lacking the supplement that allows selection of the plasmid.

Isolation of pdc1-8 suppressor mutations
Spontaneous pdc1-8 revertants/suppressors were isolated by spreading YPD plates with 1 x 10⁷ cells per plate of the pdc1-8 mutant strain YSH 4.132.-1C pregrown in YPE. Approximately 20 colonies appeared per plate after 5 days of incubation at 30 °C. Colonies of different size derived from four independent experiments were purified and examined by crossing with the pdc1-8 mutant YSH 4.116.-1D and with the pdc1 pdc5 double mutant YSH 4.136.-3D and subsequent tetrad analysis.

Cloning and sequencing of PDC1 alleles
Standard techniques for working with recombinant DNA were used [26]. The pdc1-8 and pdc1-14 alleles were cloned by constructing genomic mini-libraries in pUC19 [27] as described previously [13]. Both alleles were isolated on genomic BamHI fragments of 8 kb [13]. Smaller fragments were subcloned into M13mp18 and M13mp19 [27] and the coding region was sequenced using a T7 DNA polymerase-based sequencing kit (Pharmacia-LKB). The sequences of pdc1-8 and pdc1-14 have been deposited with the EBI database under the accession numbers X77312 and X77311, respectively.

The mutant alleles pdc1-801, pdc1-802 and pdc1-803 were isolated by PCR. Two primers complementary to sequences upstream and downstream of the PDC1 coding region were designed: primer 1 (5'-CAGTGTCTCCGACGATTTGG-3') binds at position -1110 relative to the ATG start codon and primer 2 (5'-TGGTTCCACTAATTCGTCGG-3') binds 263 bp downstream of the translational stop codon. PDC1 was amplified as a 3060-bp fragment with TUB DNA polymerase (Amersham) using 100 ng genomic DNA isolated from the respective mutants as template [28]. The reaction conditions were: 35 cycles of 1 min at 95 °C, 3 min at 60 °C and 3 min at 72 °C using the buffer supplied by the manufacturer. Initial experiments with Taq-polymerase (Promega) gave unacceptably high rates of incorrect nucleotide incorporations, whereas PCR products synthesized by TUB DNA polymerase were shown to be error-free. The PCR products were purified using the Magic^TM PCR DNA purification system from Promega, digested with PstI and Asp700 (restriction sites of the PDC1 flanking regions about 50 bp each from the primer sites) and cloned into PstI/SmaI-cleaved YCplac22 [29]. The PDC1 coding region was sequenced using pUC19 universal and reverse primers as well as primers designed according to the known sequence of PDC1 [3]. The sequences of two independent PCR reactions were determined with identical results. The sequences have been deposited with the EBI database under the accession numbers X77313-X77315.

In-vitro construction of allele pdc1-51, in which Glu51 is replaced by alanine, has been described previously [30]. All of the mutant alleles that were identified, analysed and used in this study are summarized in Table 2.

 
Plasmids and constructs
PDC1 was subcloned on a 3.7-kb PstI fragment into the YCplac22 and the YEplac112 vector [29], pdc1-Glu51Ala on a 3.0-kb SphI/BamHI fragment and the pdc1-8 and pdc1-14 alleles on 5.7-kb PstI fragments.

The same PDC1 alleles were cloned into the pYX232 vector (R&D Systems, Inc.) by digesting the vector in a first step with EcoRI; overhanging sticky ends were then filled in with Klenow enzyme. In a second step the vector was digested with BamHI removing the ATG codon of the vector. The ORFs of the alleles were amplified by PCR using two different primers, one introducing an additional BamHI site downstream of the ORF. Primer 1 (5'-CTACTCATAACCTCACGC-3') binds at position -28 relative to the ATG start codon and primer 2 (5'-cgggatcccgTAATAATTAGAGATTAAATCG-3'; lower case indicates the sequence added to generate the BamHI site) binds 1 bp downstream of the translational stop codon. The PCR was completed by filling in the overhanging ends of the product with Klenow enzyme. Subsequently the amplified fragment was digested with BamHI, resulting in one blunt and one sticky end. This fragment was ligated into the digested pYX232 vector.

For expression of pdc1-8 and pdc1-803 in Escherichia coli, the alleles were cloned into the pET22b(+):PDC1 vector [30]. The pET22b(+):pdc1-8 and pET22b(+):pdc1-803 plasmids were transformed into E. coli strain BL21 (DE3) for the expression of the recombinant Pdc proteins [31].

Northern blot analysis
Northern blot analysis was performed according to de Winde et al. [32]. The following oligonucleotide probes were used for hybridization: PDC1, 5'-ACCAAGATGGTGTCAATGACTTCCTT-3' (3'nucleotide at position +600); PDC5, 5'-GATGAATTCAACAACAGTTCTAACA-3' (starting position of the 3' nucleotide at position + 592); actin (ACT1), 5'-AATCGATTCTCAAAATGGCGTGAGTG-3' (3' nucleotide at position +469).

Western blot analysis
For Western blot analysis of crude yeast extracts [33] antiserum raised against S. cerevisiae Pdc1p [7] or against Pdc purified from baker's yeast (unpublished data) was used.

Enzyme assays in yeast crude extracts
Pdc specific activity in yeast crude extracts was determined in buffer containing 50 mM imidazole, 100 mM KCl, 10 mM MgCl₂, 0.1 mM EDTA at pH 6.8 [23]. The reaction contained 30 mM pyruvate, 0.2 mM NADH, 2 mM ThDP and 5 U·mL^-1 alcohol dehydrogenase. The oxidation of NADH was followed at 340 nm. -Galactosidase was measured as described previously [28]. The protein content of extracts was determined by the microbiuret method [34].

Reproducibility
Northern and Western blot analyses were performed at least in duplicate, and biochemical analyses at least in triplicate. Transformed yeast cells tend to give relatively high variations from experiment to experiment, and so independent experiments were performed three or four times; data are given as means and SD (Table 3). Enzymatic determination with untransformed strains was performed at least three times with SD < 20%; mean values are shown (Table 4).

 
Expression and purification of mutant Pdc
E. coli BL21 cells were grown in Luria-Bertani medium containing 100 µg·mL^-1 ampicillin, 1 mM thiamin chloride and 1 mM MgSO₄ at 37 °C with shaking. Pdc expression was induced in late log phase by the addition of 0.2 mM isopropyl--D-thiogalactopyranoside. The cells were harvested by centrifugation and resuspended in 40 mL 20 mM potassium phosphate buffer, pH 6.8, containing 1 mM Na₂EDTA, 2 mM MgSO₄, 1 mM phenylmethanesulfonyl fluoride, 1 mM ThDP, 5 mM dithiothreitol and 0.05% (w/v) reduced Triton X-100. The cell suspension was disrupted at 20 kHz in an ice-bath for 6 min on a 550 Sonic Dismembrator (Fisher Scientific) and centrifuged at 29 000 g at 4 °C for 30 min. The precipitate was discarded. Ammonium sulphate was added to the supernatant to a final concentration of 1.5 M, the solution was stirred at room temperature for 15-30 min and then centrifuged at 29 000 g at 4 °C for 15 min. Ammonium sulphate was added to the supernatant to a final concentration of 2.8 M under continuous stirring at room temperature for 15-30 min. The pellet containing the crude enzyme was collected by centrifugation at 29 000 g at 4 °C for 15 min, resuspended in 3-5 mL of 20 mM Bis/Tris (bis[2-hydroxyethyl]iminotris[hydroxymethyl]methane), pH 6.1, containing 1 mM Na₂EDTA, 2 mM MgSO₄, 0.5 mM phenylmethanesulfonyl fluoride and 1 mM ThDP and dialysed against the same buffer at 4 °C overnight. The desalted enzyme solution was loaded on a HiLoad Q Sepharose HP (column 26 x 100) equilibrated with 20 mM Bis/Tris, pH 6.1, containing 1 mM Na₂EDTA, 2 mM MgSO₄, 0.5 mM phenylmethanesulfonyl fluoride. The protein was eluted by a linear gradient with 0 to 1 M NaCl at a flow rate of 4.0 mL·min^-1. Fractions (4.0 mL·fraction^-1) were collected and evaluated for protein content and Pdc activity, and then checked for purity using SDS/PAGE. The preparation was concentrated to 10-20 mg·mL^-1 of protein and exchanged into 100 mM potassium phosphate, pH 6.1, containing 1 mM Na₂EDTA, 20 mM MgSO₄, 1 mM phenylmethanesulfonyl fluoride, 20 mM ThDP and 0.05% (w/v) NaN₃ using Amicon Centriprep 30 devices. For long-term storage, glycerol was added to the preparation to a final concentration of 30% (v/v), and the preparation was stored at -20 °C.

The purified protein was analysed by SDS/PAGE according to Laemmli [33]. The protein sample was diluted at least fivefold with sample buffer, containing 0.5 M Tris/HCl, pH 6.8, 10% (v/v) glycerol, 2% (w/v) SDS, 0.05% (v/v) 2-mercaptoethanol, and 0.00125% bromophenol blue, and heated to 100 °C for 3-5 min before loading the sample (1-5 µg protein) on the gel. The gel was stained with Coomassie Brilliant Blue G-250. Calibration proteins for SDS/PAGE (Bio-Rad) were used as molecular mass markers.

The activity of recombinant Pdc was monitored at 340 nm with the coupled enzyme assay using alcohol dehydrogenase (ADH) and NADH [35]. All components of the reaction mixture were dissolved in 100 mM Mes buffer, pH 6.0, containing 0.1% (w/v) BSA, 0.5 mM NADH, 6 U·mL^-1 ADH and 2-5 U·mL^-1 Pdc. The final concentration of pyruvate varied from 0.02 to 100 mM. The reaction was initiated by the addition of enzyme and performed at 25 °C or 30 °C. Protein was determined according to Bradford [36]. One unit of activity is defined as the amount of Pdc required to convert 1 µmol of pyruvate to acetaldehyde per min at 25 °C and pH 6.0.

 
Pdc1-8 and pdc1-14 are point mutant alleles
Schmitt and Zimmermann have previously isolated alleles of PDC1, pdc1-8 and pdc1-14, both of which conferred very low Pdc specific activity and an inability to grow on glucose medium [23]. Also in our hands, pdc1-8 and pdc1-14 strains had less than 2% residual Pdc specific activity (Table 4; unpublished data). We show here that both mutant alleles are expressed as detectable proteins as determined by Western blot analysis of extracts from mutant strains lacking the PDC5 gene (Fig. 1).

 
The pdc1-8 and pdc1-14 alleles were cloned from the genomes of the mutant strains and their entire coding regions sequenced; each mutant allele had one nucleotide change as compared with the wild-type, both leading to amino acid replacement. In pdc1-8 Asp291 was substituted for asparagine and in pdc1-14 Ser455 was replaced by phenylalanine (Table 2). Asp291 is located in a loop region in the -domain; Ser455 is located within a long helix within the -domain, which also includes the well-conserved signature motif of ThDP-dependent enzymes [17,19,22,37,38].

Effects on PDC5 expression conferred by cloned PDC1 alleles
The alleles pdc1-8 and pdc1-14, and an in vitro generated Glu51Ala allele were cloned, together with the PDC1 promoter region, into the low-copy vector YCplac22 and the multicopy vector YEplac112. The Glu51Ala mutation diminished drastically ThDP binding, and purified recombinant Glu51Ala Pdc1p had a greatly reduced (Y. Gao and F. Jordan, unpublished data) or undetectable catalytic activity [30].

All mutant alleles caused diminished Pdc specific activity when introduced into the wild-type strain (Table 3), which is consistent with the previous observation that pdc1-8 and pdc1-14 are semidominant mutations [23]. When introduced into the pdc1 strain, the pdc1-8 and pdc1-14 alleles diminished Pdc activity to as little as 10-20%, whereas pdc1-Glu51Ala caused reduction to about 50% of the wild-type activity (Table 3). As Pdc activity in the pdc1 strain is due to the expression of PDC5, the diminished Pdc activity in the transformants was likely to be the consequence of both low or absent catalytic activity of Pdc1p, and of reduced expression of PDC5. Indeed, Northern blot analysis showed that in particular the alleles pdc1-8 and pdc1-14 strongly diminished expression of PDC5 in a pdc1 background when expressed from the low-copy vector (Fig. 2). Thus, pdc1-8 and pdc1-14 appear to encode catalytically defective Pdc1p that, however, can repress the expression of PDC5. The allele pdc1--Glu51A2a caused repression of the PDC5 mRNA level only when strongly overexpressed (Fig. 2).

 
As expected, the expression of the mutant pdc1 alleles in a pdc1 strain caused a growth impairment when cells were shifted from ethanol to glucose medium as a carbon source (Fig. 3). Pdc1-8 conferred the strongest effect, already clearly apparent on the low-copy plasmid. The growth inhibition conferred by pdc1-14 and by pdc1-51 was more pronounced when the genes were transformed on a multicopy plasmid.

 
The promoter of PDC1 does not affect PDC5 expression
Next, the question of whether the promoter region of PDC1 was involved in the effect on transcription of PDC5 was addressed. Such a promoter effect could occur by competition for transcription factors binding to specific sites in the two upstream sequences. The promoter of the glycolytic gene TPI1 was fused to the coding regions of PDC1 and the pdc1 alleles and the constructs, were introduced into a pdc1 strain on the multicopy vector pYX232. The plasmid containing the wild-type PDC1 gene conferred a 10-12-fold higher Pdc specific activity, demonstrating a very high overexpression. All plasmids, except for the empty vector, strongly repressed the level of PDC5 mRNA (Fig. 2). Thus, the coding region of PDC1 wild-type and point mutant alleles is sufficient to repress PDC5.

Deletions encompassing different parts of PDC1 were then constructed and their effect on PDC5 expression tested. The different PDC1 deletion alleles were the previously described pdc1-1, which encompasses only part of the ORF and is still expressed into a shorter mRNA ([3,13] and Fig. 4A), a deletion removing the entire PDC1 ORF (pdc1-2) and a deletion of the entire ORF plus the upstream promoter region (pdc1-3). All three mutations gave approximately the same PDC5 mRNA levels (Fig. 4A). In addition, the glucose induction kinetics of Pdc5p activity in these three pdc1 strains was virtually superimposable (Fig. 4B). We conclude that the effects conferred by PDC1 on PDC5 expression are independent of the PDC1 promoter.

 
Isolation of mutations suppressing pdc1-8
To learn more about the mechanism by which PDC1 represses transcription of PDC5 we decided to search for mutations that could suppress the inability of a pdc1-8 mutant to grow on glucose. Twenty such mutants were isolated from four independent experiments. All mutations were recessive and segregated 2 : 2, indicating that a single mutation conferred the glucose-positive phenotype. In order to determine if growth restoration required PDC5, each of the 20 mutants was crossed with a pdc1 pdc5 double mutant. In 17 mutants, growth on glucose did not require PDC5 and cosegregated with PDC1. Thus, in these mutants the original pdc1-8 mutation has either reverted or was compensated by a second mutation within PDC1, resulting in a catalytically active Pdc1p. In the remaining three mutants, growth on glucose also cosegregated with PDC1 but was strictly dependent on the presence of the PDC5 gene. Hence, these mutations appeared to affect the repressive function of Pdc1p and allowed expression of PDC5.

Phenotypic characterization of pdc1-8 suppressor mutants
The mutations pdc1-801 and pdc1-802 conferred growth on glucose apparently indistinguishable from that of the wild-type and the pdc1 strain, whereas the mutant carrying allele pdc1-803 grew much more slowly on glucose, albeit better than the pdc1-8 mutant (Fig. 5). Consistent with this observation, the Pdc activity conferred by pdc1-801 and pdc1-802 was comparable to that of the pdc1 strain, whereas that of the pdc1-803 mutant was much lower (Table 4). In all three mutants Pdc activity was entirely dependent on PDC5 demonstrating that none of the novel pdc1 alleles had regained catalytic activity but rather that expression of PDC5 was stimulated (Table 4). This was confirmed by Northern blot analysis (Fig. 6) as well as by determination of -galactosidase activity as a measure of PDC5 promoter activity using a PDC5-lacZ promoter-reporter construct (Table 4). The PDC5 mRNA level in mutants pdc1-801 and pdc1-802 was similar to that in the pdc1 mutant, and that in the pdc1-803 mutant was considerably lower (Fig. 6). Consistent with this, the PDC5 promoter activity in mutants pdc1-801 and pdc1-802 was almost as high as in a pdc1, strain whereas in the pdc1-803 mutant it was only approximately twice of that of the wild-type (Table 4). In that mutant, the level of the PDC1 mRNA was consistently found to be increased (Fig. 6).

 
We have shown previously that expression of PDC5 in medium with ethanol as a carbon source is essentially undetectable, even in a pdc1 strain [3,5]. Moreover, no PDC5 mRNA could be detected when any of the mutants studied here was grown on ethanol medium (Fig. 6), and there was also only extremely low -galactosidase activity produced from the PDC5-lacZ reporter construct (data not shown). This observation is consistent with an independent control of PDC5 expression by Pdc1p and by the carbon source.

Pdc1-801 and pdc1-802 are nonsense alleles, pdc1-803 has a deletion of two codons
The novel mutant alleles were cloned by PCR and the sequence was determined (Table 2). The mutations pdc1-801 and pdc1-802 caused a premature translation stop. In allele pdc1-801, the C in position 1091 of the coding region was deleted which leads to translational stop at codon 365. Pdc1-802 is characterized by replacement of a C at position 751 with a T which changes this codon (239) into a stop codon. There was a second mutation in pdc1-801 that caused an amino acid replacement (Table 2). In Western blot analysis using antibodies raised against Pdc1p no protein product could be detected from these alleles, neither full length Pdc1p (Fig. 7) nor a truncated product (data not shown).

 
In allele pdc1-803, six nucleotides were missing leading to the deletion of codons 295 and 296, just four codons downstream of the original pdc1-8 mutation (Table 2). Pdc1-803 was expressed into a detectable protein product (Fig. 7).

Expression and purification of Pdc from pdc1-8 and pdc1-803
The pdc1-8 and pdc1-803 mutant genes were expressed in E. coli BL21 as described previously for wild-type PDC1 [6]. The expressed proteins were found in the soluble fraction after cell lysis. The specific activity for the pdc1-8 and pdc1-803 variants in the crude extract was at least 40-50 times lower than that of the wild-type Pdc (Table 5). The two variants were successfully purified to homogeneity and exhibited a single band of M_r 60 kDa as judged by SDS/PAGE (Fig. 8). While there was evidence for activity in the crude extracts of the variants over and above background, on purification to homogeneity both the pdc1-8 and pdc1-803 variants showed very low activity. ThDP was separated from the protein by the ion-exchange column at pH 6.1, and, as a result, both variants were purified as apo-enzymes. Even after 2 h incubation at room temperature with 20 mM ThDP and 20 mM MgSO₄ in 100 mM phosphate (pH 6.0), the activity remained very low (Table 5).

 
Previous work has shown that expression of the genes PDC1 and PDC5, which encode the isoforms of yeast Pdc, is controlled by an autoregulation mechanism [2,3,11,13]. This autoregulation first became apparent - and is most conveniently studied - by the transcriptional induction of PDC5 expression in a pdc1 mutant. In this work we provide strong evidence that the effect conferred by Pdc1p on expression of PDC5 is independent of the enzymatic activity of Pdc1p but is due rather to an as yet unknown regulatory role of this enzyme.

Control of PDC5 expression by Pdc1p does not require Pdc activity
The most obvious explanation for the stimulation of PDC5 (and PDC1) promoter activity by deletion of the PDC1 gene had been that the accumulation of pyruvate or other glycolytic metabolites triggers such an effect. In fact, evidence has been provided previously that the stimulation of expression of PDC1 by glucose correlates with the level of metabolites in the lower part of the glycolytic pathway, i.e. triosephosphates [39]. Moreover, Liesen et al. [11] described a promoter element in the PDC1 upstream region, which appeared to be responsible both for carbon source control and autoregulation. Our data are inconsistent with such an overlap between autoregulation and carbon source control. Firstly, pdc1-8 mutants have been reported to accumulate and secrete large amounts of pyruvate into the growth medium [23] and probably also accumulate triosephosphates, i.e. metabolites upstream of pyruvate in glycolysis. However, as shown here, neither expression of the PDC1 gene nor that of PDC5 is stimulated at all in a pdc1-8 mutant (Fig. 6). Moreover, the stimulation of PDC5 expression by deletion of PDC1 is only apparent in glucose medium, i.e. expression of PDC5 is strongly controlled by the carbon source even in a pdc1 strain (Fig. 6) [3,5,12] clearly separating carbon source control from autoregulation, at least for PDC5 (the promoter of PDC1 does not appear to respond anymore to the carbon source in a pdc1 mutant [11]). Simple marker effects for the observed stimulation of PDC5 expression in pdc1 mutants can also be excluded. The LEU2 marker gene used in this study has been shown to affect pyruvate metabolism [15]. However, the deletion of PDC1 has been performed previously with different markers giving the same effect on PDC5 expression [2]. Thus, we conclude that the autoregulatory effect on PDC gene expression is independent of both sugar catabolism and the catalytic activity of Pdc1p.

Autoregulation requires Pdc1p
Although the catalytic activity of Pdc1p does not seem to be required for autoregulation, i.e. for repression of PDC5 (and of PDC1), it does seem that a property of the protein mediates the transcriptional effect. Because any of the PDC1 alleles studied in this work can confer repression of PDC5, even when expressed under the control of a different promoter (Fig. 2), we can exclude simple competition for transcription factors binding upstream sequences as being involved in the effect. This is further confirmed by the finding that deletion of parts of the PDC1 ORF, the entire ORF and the ORF plus the promoter cause exactly the same stimulation of PDC5 expression. Finally, the activity of the promoter of PDC1 is strongly stimulated when the ORF of PDC1 is partially or completely deleted suggesting a mechanism that affects both promoters at the same time and not in competition (Fig. 4) [3,11,13].

Some property of the PDC1 mRNA can also be excluded with high probability as a mediator of the response. The alleles pdc1-801 and pdc1-802 are both transcribed into mRNA but are not expressed as protein and lead to high level PDC5 expression. On the other hand, the mutants pdc1-8 and pdc1-14 are expressed as protein and mediate repression of PDC5. Thus mediation of the autoregulatory effect seems to be a function of Pdc1p and not its mRNA.

Which property of Pdc1p could mediate autoregulation?
We cannot exclude the possibility that Pdc1p is partly localized to the nucleus and controls transcription directly. However, we regard this possibility as extremely unlikely as S. cerevisiae Pdc is highly homologous to those from other yeasts and bacteria over their entire sequence, and yeast Pdc in particular does not show any specific sequence features that would point to a nuclear localization, DNA binding or transcriptional regulation. Rather, we favour the idea that Pdc1p either interacts directly with other regulatory proteins or has another activity that signals to the transcriptional machinery. Hence, the question appears to be what kind of signal does Pdc1p produce?

There are numerous studies on the ability of yeast Pdc to perform reactions other than decarboxylation of pyruvate. Pdc and related enzymes can be used for chemoenzymatic syntheses and biotransformations [40,41]. Those reactions are of two types: either substrates analogous to pyruvate can be decarboxylated or the activated aldehyde bound to the enzyme as an intermediate can be transferred to another aldehyde instead of water in a benzoin-type condensation, yielding for example acetoin from pyruvate plus acetaldehyde. The latter reaction may even occur at a low level in intact yeast cells as it involves both product and substrate of the enzyme. Pdc has also been thought to be involved in the natural production of fusel alcohols by yeast cells starting from branched chain amino acids [42]. However, all of these activities are highly likely to require the same binding sites and catalytic mechanism, as does the decarboxylation of pyruvate, and hence would be expected to be defective in mutants such as pdc1-8 and pdc1-14 as well; therefore we believe it unlikely that any alternative product produced by Pdc could trigger the transcriptional control mechanism.

In addition to binding its substrate, pyruvate, Pdc also binds two cofactors, ThDP and Mg²⁺ [43]. One could speculate that the association with those cofactors or their level could play a role in the regulatory function of Pdc1p. As Mg²⁺ ions are required in many other reactions, we regard their involvement specifically in PDC gene regulation as rather unlikely. ThDP, however, is only used by a total of five well-characterized enzymes in yeast metabolism, plus three proteins of unknown function identified by systematic sequencing [16]. Remarkably, recent evidence shows that expression of PDC5 is indeed controlled by the level of thiamin in the growth medium, i.e. it is repressed by high thiamin levels. However, we regard it as rather unlikely that a diminished ability for ThDP binding by Pdc1p provides the relevant signal for autoregulation. Firstly, deletion of PDC1 would be expected to result in a higher ThDP level and hence in repression of PDC5 expression; however, expression of PDC5 is stimulated in the pdc1 strain. Secondly, the pdc1-Glu51Ala allele, which was chosen because of its demonstrated defect in ThDP binding [30], can still repress PDC5, albeit only partially. Thirdly, whereas autoregulation affects the promoter activity of both PDC1 and PDC5 the effect of thiamin is restricted to PDC5 [16a]. However, we do not wish to exclude entirely at this stage the possibility that a complex regulatory interplay between thiamin metabolism and the production of ThDP-dependent enzymes, which seems to exist [16], might be involved in the autoregulatory effect.

The phenotype of the pdc1-803 allele appears to provide evidence that Pdc autoregulation can affect PDC1 and PDC5 expression differently. In that mutant, expression of PDC5 is stimulated only moderately whereas that of PDC1 seems to be strongly enhanced. In contrast, in the pdc1-801 and pdc1-802 mutants it seems that only the PDC5 mRNA level is strongly increased. However, it is well known that nonsense mutations cause mRNA instability [44] and hence the apparently normal levels of PDC1 mRNA in pdc1-801 and pdc1-802 could be the result of balancing effects of enhanced transcription and mRNA degradation.

Effects of the mutations on structure and function of Pdc1p
The structure of Pdc has been determined to 2.3 Å and 2.7 Å resolution from crystals produced in the absence [38,45] and in the presence of the substrate analogue pyruvamide [22], respectively. In addition, a large number of genes encoding Pdc and the closely related acetolactate synthases have been cloned and sequenced. This allows for some interpretation and speculation on the effects of the mutations analysed here.

Asp291Asn in pdc1-8 lies within a loop region (residues 292-303) between two helices in the domain (Fig. 9). This loop region appears to be very flexible because it could only be resolved in two of the four subunits from crystals grown in the presence of pyruvamide [17,19,22,38,45]. The amino acids in this loop are not very well conserved among different Pdc with the notable exception of Asp291. There is an aspartic acid in this position in 19 out of 21 Pdc sequences inspected and the residues immediately flanking this aspartic acid seem to be conserved with respect to both size and functionality. Thus, Asp291 may have specific importance. The observation, that the Asp291Asn mutation as well as the additional deletion of amino acids 295 and 296 causes the enzyme to loose ThDP irreversibly during purification, suggests that this loop has structural significance. Certainly more work is required to understand the role of this loop in the structure and function of Pdc.

 
Ser455Phe in pdc1-14 lies within a long helix in the domain within the well-conserved signature sequence of ThDP-dependent enzymes. Ser455 is not particularly well conserved but in none of the more than 50 ThDP-dependent enzymes inspected was the residue in this position larger than glutamic acid. Thus, phenylalanine in this position may distort or bend the helix and hence affect co-ordination of the Mg²⁺ ion, which is mediated by residues in the neighbourhood [17,19]. Previous site-directed mutagenesis on pyruvate dehydrogenase has shown that mutations in this so-called ThDP fold have dramatic effects that can be attributed to distortion of the Mg²⁺ co-ordination sphere [18,46].

Apparently, the products of pdc1-8 and pdc1-14 must have some property that allows recognition by the autoregulatory system controlling PDC gene expression. It is possible that the proteins derived from the mutant genes permanently adopt a conformation that resembles that conformation of Pdc1p, which mediates the transcriptional effect. According to such a scenario, the deletion of two additional amino acids in pdc1-803 may partially compensate for the defects, so that the protein can adopt the conformation required for transcriptional activation.

Our results suggest strategies for additional experiments to elucidate further the role of Pdc1p in the control of expression of the yeast PDC genes by this unusual regulatory system.

 
The authors thank Profs F. K. Zimmermann (Darmstadt) and J. M. Thevelein (Leuven) in whose laboratories this work was conducted. We thank R. Bill (Göteborg) for critical reading of the manuscript. Part of this work was supported by the Commission of the European Union via contract BIO4-CT95-0161 to J. M. Thevelein and S.H. The Jordan lab was supported by grant USPHS-NIH-GM-50380 from the Rutgers Busch Fund. Collaboration between the Hohmann and the Jordan (and W. Furey) groups was supported by NATO travel grant CRG.951237. Work in the König group was supported by grants 05-641KEBO and 03-K04HAL-2 from the German Federal Ministry for Education and Research (BMBF).

 
Enzymes: pyruvate decarboxylase (EC4.1.1.1).

Note: Ines Eberhardt is now at the Departement Moleculaire Biologie, Eenheid voor toegepaste en fundamentele Biologie, Universiteit Gent, Belgium; Håkan Cederberg is now at the Department of Genetic and Cellular Toxicology, Wallenberg Laboratory, Stockholm University, Sweden.

1. Schmitt, H.D., Ciriacy, M. & Zimmermann, F.K. ( 1983) The synthesis of yeast pyruvate decarboxylase is regulated by large variations in the messenger RNA level. Mol. Gen. Genet. 192, 247- 252.
2. Seeboth, P.G., Bohnsack, K. & Hollenberg, C.P. ( 1990) pdc1° mutants of Saccharomyces cerevisiae give evidence for an additional structural PDC gene: cloning of PDC5, a gene homologous to PDC1. J. Bacteriol. 172, 678- 685.
3. Hohmann, S. & Cederberg, H. ( 1990) Autoregulation may control the expression of yeast pyruvate decarboxylase structural genes PDC1 and PDC5. Eur. J. Biochem. 188, 615- 621.
4. Hohmann, S. ( 1997) Pyruvate Decarboxylases. In Yeast Sugar Metabolism. (Zimmermann, F.K. & Entian, K.-D., eds), pp. 187- 211. Technomic Publishing Co, Inc., Lancaster.
5. Hohmann, S. ( 1991) Characterization of PDC6, a third structural gene for pyruvate decarboxylase in Saccharomyces cerevisiae. J. Bacteriol. 173, 7963- 7969.
6. Baburina, I., Gao, Y., Hu, Z., Jordan, F., Hohmann, S. & Furey, W. ( 1994) Substrate activation of brewer's yeast pyruvate decarboxylase is abolished by mutation of cysteine 221 to serine. Biochem. J. 33, 5630- 5635.
7. Killenberg-Jabs, M., König, S., Hohmann, S. & Hübner, G. ( 1996) Purification and characterisation of the pyruvate decarboxylase from a haploid strain of Saccharomyces cerevisiae. Biol. Chem. Hoppe-Seyler 377, 313- 317.
8. Hohmann, S. ( 1991) PDC6, a weakly expressed pyruvate decarboxylase gene from yeast, is activated when fused spontaneously under the control of the PDC1 promoter. Curr. Genet. 20, 373- 378.
9. Zeng, X., Farrenkopf, B., Hohmann, S., Dyda, F., Furey, W. & Jordan, F. ( 1993) Role of cysteines in the activity and inactivation of brewers's yeast pyruvate decarboxylase investigated with a PDC1-PDC6 fusion protein. Biochemisitry 32, 2704- 2709.
10. Butler, G. & McConnell, D.J. ( 1988) Identification of an upstream activation site in the pyruvate decarboxylase structural gene (PDC1) of Saccharomyces cerevisiae. Curr. Genet. 14, 405- 412.
11. Liesen, T., Hollenberg, C.P. & Heinisch, J.J. ( 1996) ERA, a novel cis-acting element required for autoregulation and ethanol repression of PDC1 transcription in Saccharomyces cerevisiae. Mol. Microbiol. 21, 621- 632.
12. Hohmann, S. ( 1993) Characterisation of PDC2, a gene necessary for high level expression of pyruvate decarboxylase structural genes in Saccharomyces cerevisiae. Mol. Gen. Genet. 241, 657- 666.
13. Schaaff, I., Green, J.B.A., Gozalbo, D. & Hohmann, S. ( 1989) A deletion of the PDC1 gene for pyruvate decarboxylase of yeast causes a different phenotype than previously isolated point mutations. Curr. Genet. 15, 75- 81.
14. Flikweert, M.T., Vanderzanden, L., Janssen, W., Steensma, H.Y., van Dijken, J.P. & Pronk, J.T. ( 1996) Pyruvate decarboxylase: an indispensable enzyme for growth of Saccharomyces cerevisiae on glucose. Yeast 12, 247- 257.
15. Pronk, J.T., Steensma, H.Y. & van Dijken, J.P. ( 1996) Pyruvate metabolism in Saccharomyces cerevisiae. Yeast 12, 1607- 1633.
16. Hohmann, S. & Meacock, P.A. ( 1998) Thiamin metabolism and thiamindiphosphate-dependent enzymes in the yeast Saccharomyces cerevisiae: genetic regulation. Biochim. Biophys. Acta 1385, 201- 219.
16. Muller, E.H., Richards, E., Norbeck, J., Byrne, K.L., Karlsson, K.A., Pretorius, G.H.J., Meacock, P.A., Blomberg, A & Hohmann, S. ( 1999) Thiamin repression and pyruvate decarboxylase autoregulation independently control the expression of the Saccharomyces cerevisiae PDC5 gene. FEBS Lett. in press.
17. Furey, W., Arjunan, P., Chen, L., Sax, M., Guo, F. & Jordan, F. ( 1998) Structure-function relationship and flexible tetramer assembly in pyruvate decarboxylase revealed by analysis of crystal structures. Biochim. Biophys. Acta 1385, 253- 270.
18. Jordan, F., Nemeria, N., Guo, F., Baburina, I., Gao, Y., Kahyaoglu, A., Li, H., Wang, J., Yi, J., Guest, J.R. & Furey, W. ( 1998) Regulation of thiamin diphosphate-dependent 2-oxo acid decarboxylases by substrate and thiamin diphosphate. Mg (II) - evidence for tertiary and quaternary interactions. Biochim. Biophys. Acta 1385, 287- 306.
19. König, S. ( 1998) Subunit structure, function and organisation of pyruvate decarboxylase from various organisms. Biochim. Biophys. Acta 1385, 271- 286.
20. Baburina, I., Dikdan, G., Guo, F., Tous, G.I., Root, B. & Jordan, F. ( 1998) Reactivity at the substrate activation site of yeast pyruvate decarboxylase: inhibition by distortion of domain interactions. Biochemistry 37, 1245- 1255.
21. Baburina, I., Li, H., Bennion, B., Furey, W. & Jordan, F. ( 1998) Interdomain information transfer during substrate activation of yeast pyruvate decarboxylase: the interaction between cysteine 221 and histidine 92. Biochemistry 37, 1235- 1244.
22. Lu, G., Dobritzsch, D., König, S. & Schneider, G. ( 1997) Novel tetramer assembly of pyruvate decarboxylase from brewer's yeast observed in a new crystal form. FEBS Lett. 403, 249- 253.
23. Schmitt, H.D. & Zimmermann, F.K. ( 1982) Genetic analysis of the pyruvate decarboxylase reaction in yeast glycolysis. J. Bacteriol. 151, 1146- 1152.
24. Eberhardt, I. & Hohmann, S. ( 1995) Strategy for deletion of complete open reading frames in Saccharomyces cerevisiae. Curr. Genet. 27, 306- 308.
25. Sherman, F., Fink, G.R. & Hicks, J.B. ( 1983) Methods in Yeast Genetics. Cold Spring Harbor Laboratory Press, Cold Spring, Harbor, NY.
26. Sambrook, J., Fritsch, E.F. & Maniatis, T. ( 1989) Molecular cloning. A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring, Harbor, NY.
27. Yanisch-Perron, C., Vieira, J. & Messing, J. ( 1985) Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13mp18 and pUC19 vectors. Gene 33, 103- 119.
28. Rose, M.D., Winston, F. & Hieter, P. ( 1990) Methods in yeast genetics. A Laboratory Course Manual. Cold Spring Harbor Laboratory Press, Cold Spring, Harbor, New York.
29. Gietz, R.D. & Sugino, A. ( 1988) New yeast-E. coli shuttle vectors with in vitro mutagenized yeast genes lacking six base pair restriction sites. Gene 74, 527- 534.
30. Killenberg-Jabs, M., König, S., Eberhardt, I., Hohmann, S. & Hübner, G. ( 1997) Role of Glu51 for cofactor binding and catalytic activity in pyruvate decarboxylase from yeast studied by site directed mutagenesis. Biochemistry 36, 1900- 1905.
31. Studier, F.W. & Moffatt, B.A. ( 1986) Use of bacteriophage T7 RNA polymerase to direct selective high-level expression of cloned genes. J. Mol. Biol. 189, 113- 130.
32. de Winde, J.H., Crauwels, M., Hohmann, S., Thevelein, J.M. & Winderickx, J. ( 1996) Differential requirement of the yeast sugar kinases for sugar sensing in the establishment of the catabolite repressed state. Eur. J. Biochem. 241, 633- 643.
33. Laemmli, U.K. ( 1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 277, 680- 685.
34. Zamenhoff, S. ( 1957) Preparation and assay of deoxyribonucleic acids from animal tissues. Methods Enzymol. 3, 696- 704.
35. Holzer, H., Schultz, G., Villar-Palasi, C. & Jüntgen-Sell, J. ( 1956) Isolierung der Hefecarboxylase und Untersuchung über die Aktivität des Enzyms in lebenden Zellen. Biochem. Z. 327, 331- 344.
36. Bradford, M.M. ( 1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248- 254.
37. Muller, Y.A., Lindqvist, Y., Furey, W., Schulz, G.E., Jordan, F. & Schneider, G. ( 1993) A thiamin diphosphate binding fold revealed by comparison of the crystal structures of transketolase, pyruvate oxidase and pyruvate decarboxylase. Structure 1, 95- 103.
38. Arjunan, P., Umland, T., Dyda, F., Swaminathan, S., Furey, W., Sax, M., Farrenkopf, B., Gao, Y., Zhang, D. & Jordan, F. ( 1996) Crystal structure of the thiamin diphosphate-dependent enzyme pyruvate decarboxylase from the yeast Saccharomyces cerevisiae at 2.3 Ångstrom resolution. J. Mol. Biol. 256, 590- 600.
39. Boles, E. & Zimmermann, F.K. ( 1993) Induction of pyruvate decarboxylase in glycolysis mutants of Saccharomyces cerevisiae correlates with the concentrations of three-carbon glycolytic metabolites. Arch. Microbiol. 160, 324- 328.
40. Iding, H., Siegert, P., Mesch, K. & Pohl, M. ( 1998) Application of alpha-keto acid decarboxylases in biotransformations. Biochim. Biophys. Acta 1385, 307- 322.
41. Schörken, U. & Sprenger, G.A. ( 1998) Thiamin-dependent enzymes as catalysts in chemoenzymatic syntheses. Biochim. Biophys. Acta 1385, 229- 243.
42. ter Schure, E.G., Flikweert, M.T., van Dijken, J.P., Pronk, J.T. & Verrips, C.T. ( 1998) Pyruvate decarboxylase catalyzes decarboxylation of branched-chain 2-oxo acids but it is not essential for fusel alcohol production in Saccharomyces cerevisiae. Appl. Environ. Microbiol. 64, 1303- 1307.
43. Schellenberger, A. & Hübner, G. ( 1967) Zur Theorie der Thiaminpyrophosphat Wirkung: IV. Mechanismus und Kinetik der Rekombination und daraus abgeleitete Bindungsverhältnisse im aktiven Zentrum der Hefe-PDC. Hoppe-S. Z. Physiol. Chem. 348, 491- 500.
44. Peltz, S.W., He, F., Welch, E. & Jacobson, A. ( 1994) Nonsense-mediated mRNA decay in yeast. Prog. Nucl Acid Res. Mol. Biol. 47, 271- 298.
45. Dyda, F., Furey, W., Subramanyam, S., Sax, M., Farrenkopf, B. & Jordan, F. ( 1993) Catalytic centers of the thiamin diphosphate dependent enzyme pyruvate decarboxylase at 2.4 Å resolution, Biochemistry 32, 6165- 6170.
46. Yi, J., Nemeria, N., McNally, A., Jordan, F., Machado, R.S. & Guest, J.R. ( 1996) Effect of substitutions in the thiamin diphosphate-magnesium fold on the activation of the pyruvate dehydrogenase complex from Escherichia coli by cofactors and substrate. J. Biol. Chem. 271, 33192- 33200.

(Received 16 November 1998; revised 23 February 1999; accepted 1 March 1999)
 

(Full Text online)

© 2005 Transgalactic Ltd (manufacturer of Bioscreen C software) | Privacy Statement | P.O. Box 1393, 00101 Helsinki, Finland, phone: +358 9 85172920, fax: +358 9 8749481, e-mail: microbiology@bionewsonline.com
 

Last modified: May 25, 2005